Re: [gmx-users] difference of calculation time between the numbers of fep-lambdas values

2015-04-27 Thread nao.morishita
Hi Justin and Hannes,



Thank you for the replay.

I apologize for the insufficient information.

Actually, I calculated them with couple-moltype = solute and init-lambda-state 
= 1.

The other parameters except for fep-lambdas, the number of core(8) and hardware 
were same.

When I look at "REAL CCLE AND TIME ACCOUNTING" in log file, there is large 
difference in "PME redist. X/F".



What makes it much longer?

I would appreciate any comments or concerns.



Input1(fep-lambdas   = 0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0)





Computing: Nodes   Th. Count  Wall t (s) G-Cycles   %

-

Domain decomp. 81 25 100.985 2100.559 0.2

DD comm. load  81501   0.0040.082 0.0

Neighbor search81 250001 351.227 7305.783 0.8

Comm. coord.   81251  74.248 1544.405 0.2

Force  812518671.720   180378.28820.8

Wait + Comm. F 81251  49.926 1038.492 0.1

PME mesh   81251   31496.259   655145.85575.6

Write traj.81547   0.2755.710 0.0

Update 81251 284.487 5917.540 0.7

Constraints81502 396.070 8238.552 1.0

Comm. energies 81 250001   7.876  163.837 0.0

Rest   8 250.084 5201.928 0.6

-

Total  8   41683.159   867041.031   100.0

-

-

PME redist. X/F81753   26513.015   551490.63263.6

PME spread/gather  81   10043879.93680705.579 9.3

PME 3D-FFT 81   1004 702.65614615.765 1.7

PME 3D-FFT Comm.   81   2008 264.536 5502.543 0.6

PME solve  81502 127.236 2646.610 0.3

-



   Core t (s)   Wall t (s)(%)

   Time:   332678.40041683.159  798.1

 11h34:43

 (ns/day)(hour/ns)

Performance:   10.3642.316

Finished mdrun on node 0 Thu Apr  2 06:31:45 2015

*

Input2(fep-lambdas   = 0.0 0.1 0.2)

*

Computing: Nodes   Th. Count  Wall t (s) G-Cycles   %

-

Domain decomp. 81 25  85.779 1784.259 0.5

DD comm. load  81501   0.0030.072 0.0

Neighbor search81 250001 312.515 6500.517 1.7

Comm. coord.   81251  69.584 1447.387 0.4

Force  812514219.96387778.08923.5

Wait + Comm. F 81251  46.401  965.178 0.3

PME mesh   81251   12342.326   256728.71768.8

Write traj.81520   0.1503.117 0.0

Update 81251 262.237 5454.706 1.5

Constraints81502 368.111 7656.961 2.1

Comm. energies 81 250001   6.524  135.712 0.0

Rest   8 227.039 4722.575 1.3

-

Total  8   17940.633   373177.290   100.0

-

-

PME redist. X/F817537982.056   166032.15844.5

PME spread/gather  81   10043327.97569224.12918.5

PME 3D-FFT 81   1004 645.03613417.186 3.6

PME 3D-FFT Comm.   81   2008 259.876 5405.587 1.4

PME solve  81502 118.906 2473.318 0.7

-



   Core t (s)   Wall t (s)(%)

   Time:   143196.26017940.633  798.2

 4h59:00

 (ns/day)(hour/ns)

Performance:   24.0790.997

Finished mdrun on node 0 

[gmx-users] difference of calculation time between the numbers of fep-lambdas values

2015-04-26 Thread nao.morishita
Dear Gromacs Users,

I would appreciate you to give me any comments or hints about the question 
below.
What makes the calculation time much longer when I calculate a solvation free 
energy using the mdp input1 compared to the mdp input2?

Input1
Free_energy   = yes
Couple-lambda0= vdw-q
Couple-lambda1= none
Couple-intramol   = no
Init-lambda-state = 2
Fep-lambdas   = 0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0
Calc-lambda-neighbors = 1
Input2
Free_energy   = yes
Couple-lambda0= vdw-q
Couple-lambda1= none
Couple-intramol   = no
Init-lambda-state = 2
Fep-lambdas   = 0.0 0.1 0.2
Calc-lambda-neighbors = 1

It takes three times longer when I write down 36 lambda values compared to 3 
lambda values even if there is no difference between mdp inputs other than the 
value of fep-lambdas and it is a normal free energy calculation not like 
Hamiltonian replica exchange calculation.

Thanks in advance and best regards,

Nao

-- 
Gromacs Users mailing list

* Please search the archive at 
http://www.gromacs.org/Support/Mailing_Lists/GMX-Users_List before posting!

* Can't post? Read http://www.gromacs.org/Support/Mailing_Lists

* For (un)subscribe requests visit
https://maillist.sys.kth.se/mailman/listinfo/gromacs.org_gmx-users or send a 
mail to gmx-users-requ...@gromacs.org.